AI coding benchmark MirrorCode published its full results June 26, showing Claude Opus 4.7 autonomously rebuilt a 60,000-line interpreter and scored 56% overall — completing tasks that take human ...
Authors Table of Contents Media Contact For general and media inquiries and to book our experts, please contact: ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results